Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

sudo make train #34

Open
oraphateixeira opened this issue Jun 9, 2021 · 1 comment
Open

sudo make train #34

oraphateixeira opened this issue Jun 9, 2021 · 1 comment
Labels
bug Something isn't working help wanted Extra attention is needed

Comments

@oraphateixeira
Copy link

Captura de tela de 2021-06-09 20-08-11
mkdir -p bot/models
docker-compose up --build coach
Building coach
Sending build context to Docker daemon 5.192MB
Step 1/5 : FROM botrequirements
---> 0810ffb2e57f
Step 2/5 : WORKDIR /bot
---> Using cache
---> 68015aa86c40
Step 3/5 : COPY ./bot /bot
---> Using cache
---> aae9d1ea02a8
Step 4/5 : COPY ./modules /modules
---> Using cache
---> 3ac6a93d786f
Step 5/5 : RUN find . | grep -E "(pycache|.pyc|.pyo$)" | xargs rm -rf
---> Using cache
---> a401317ad89a
Successfully built a401317ad89a
Successfully tagged bot-da-boss_coach:latest
Starting bot-da-boss_coach_1 ... done
Attaching to bot-da-boss_coach_1
coach_1 | rasa train -vv
coach_1 | make: *** [Makefile:21: train] Illegal instruction (core dumped)
bot-da-boss_coach_1 exited with code 2

@clarissalimab clarissalimab added bug Something isn't working help wanted Extra attention is needed labels Jun 22, 2021
@clarissalimab
Copy link

clarissalimab commented Jun 22, 2021

Atualizando a issue:

A @BrunaNayara e o @oraphateixeira descobriram que o problema é de compatibilidade com a CPU do Yan, e possivelmente outras pessoas terão problemas com isso no futuro. Uma das soluções seria fazer o downgrade do tensorflow para a versão 0.15, mas isso não é muito viável de ser feito considerando que a gente não controla diretamente essa dependência, que é do framework que estamos usando.

A solução que a Bruna propôs é criarmos uma máquina remota que tenha uma CPU compatível para realizar os treinamentos. Eu acho interessante para exercitarmos inclusive um futuro deploy do bot, que não dependerá do treinamento local para podermos disponibilizar o modelo =)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working help wanted Extra attention is needed
Projects
None yet
Development

No branches or pull requests

2 participants